Project Summary/Abstract Retinal degenerative diseases lead to blindness due to loss of photoreceptors, while neurons in the inner retinal layers are preserved to a large extent. Electronic retinal prostheses seek to reintroduce information into the visual system and thereby restore sight by electrical stimulation of surviving neurons. Clinical results with the first retinal implants demonstrated feasibility of prosthetic vision in patients blinded by retinal degeneration. However, current prostheses provide very low resolution, and being powered through inductive coils, require very complex surgical methods to implant the power supply connected to retinal stimulating array via trans-scleral cable. We developed a photovoltaic subretinal prosthesis, in which silicon photodiodes in each pixel directly convert pulsed near-infrared images projected from video goggles into local electric currents to stimulate the nearby neurons. This system offers multiple advantages over other designs: (1) Wireless system is scalable to thousands of pixels in the implant; (2) Modular design greatly simplifies surgery, allows tiling to match the eye curvature and to expand visual field; (3) Projection of stimulating patterns onto the retina maintains natural link between eye movements and image perception; (4) Network-mediated stimulation retains several important features of the retinal signal processing, including flicker fusion, adaptation to static images and non-linear summation of subunits in receptive fields, which enables high spatial resolution. We demonstrated that with pixel sizes down to 70 m, prosthetic visual acuity in rodent models of retinal degeneration matches the pixel pitch, corresponding to about 20/250 acuity in the human eye. The implants are well tolerated in the subretinal space, and responses are stable during the lifetime of the animals (12- month follow-up). While this system is being transitioning into clinical trials, we propose to double the resolution, study retinal changes under chronic stimulation, and explore intracellular connectivity. In particular, we will develop photovoltaic arrays with higher pixel density. To improve proximity and increase electrode surface area, we will integrate pillar electrodes with photovoltaic pixels. We will explore the changes in retinal wiring during the degeneration, and the effect of chronic stimulation on retinal plasticity. In addition, we will explore feasibility of the cell-attached integration of nanoelectrodes. If successful, they might enable greatly reduced stimulation thresholds ? down to the ambient light levels, and much more natural introduction of the visual information, including excitatory and inhibitory inputs mimicking the ON and OFF retinal pathways.